Malayic and the Mekong-Mamberamo Linguistic Area David Gil Gil (to appear) demonstrates the existence of a Mekong-Mamberamo linguistic area consisting of Mainland Southeast Asia, the Indonesian archipelago and western New Guinea, and characterized by the following 17 properties: (1) 17 Mekong-Mamberamo Properties 1. passing gesture 2. repeated dental clicks expressing amazement 3. conventionalized greeting with 'where' 4. 'eye day' > 'sun' lexicalization 5. d/t place-of-articulation asymmetry 6. numeral classifiers 7. verby adjectives 8. basic SVO word order 9. iamitive perfects 10. 'give' causatives 11. low differentiation of adnominal attributive constructions 12. weakly developed grammatical voice 13. isolating word structure 14. short words 15. low grammatical-morpheme density 16. optional thematic-role flagging 17. optional TAM marking Part 1 of this paper examines each of the above 17 Mekong-Mamberamo properties in turn, showing that it is typically exhibited by Malayic languages. With its heartland in Sumatra and Borneo, and its outliers extending from central Thailand to western New Guinea, Malayic is thus the Mekong-Mamberamo language family par excellence. Part 2 of this paper addresses the question how Malayic came to exhibit its present- day array of Mekong-Mamberamo properties. Given that Taiwan, the homeland of Austronesian, and the Philippines, into which the Austronesian languages first spread, both lie outside the Mekong-Mamberamo area, it is clear that when Austronesian languages spread south from the Philippines into the Indonesian archipelago, one or more languages ancestral to Malayic must have taken on the Mekong-Mamberamo properties through contact with the non-Austronesian languages, now long extinct, that they encountered in the region. What is less clear, however, is what the specific mechanisms of contact might have been that led to the Malayic languages exhibiting their contemporary linguistic profile. Of the 17 Mekong-Mamberamo properties in (1), no's 11-17 represent a state of affairs involving lesser complexity than their non-Mekong-Mamberamo counterparts. With respect to these 7 properties, then, Malayic could in principle represent the outcome of a process of contact-induced simplification; however, such an process cannot account for the remaining 10 properties, no's 1-10. In general, two competing models of language contact and convergence have been proposed that might underlie the kind of language change giving rise to the Mekong-Mamberamo nature of the Malayic family: metatypy 1 and relexification. At present, though, it is hard to adjudicate between the two models. One commonly invoked argument against relexification is that Malayic languages possess at least some morphological markers that appear to be inherited from proto- Austronesian; however, each and every one of the affixes reconstructable to proto- Malayic is also attested as having been borrowed into other, non-Austronesian languages, such as the Aslian languages of the Malay peninsula and the Papuan languages of the New Guinea Bird's Head, thereby suggesting that a similar process of borrowing into Malayic cannot be ruled out a priori. 2 .